draw: rename draw_vs_varient.c to draw_vs_variant.c
authorBrian Paul <brianp@vmware.com>
Tue, 7 Jun 2011 13:55:06 +0000 (07:55 -0600)
committerBrian Paul <brianp@vmware.com>
Tue, 7 Jun 2011 13:57:26 +0000 (07:57 -0600)
src/gallium/auxiliary/Makefile
src/gallium/auxiliary/SConscript
src/gallium/auxiliary/draw/draw_vs_variant.c [new file with mode: 0644]
src/gallium/auxiliary/draw/draw_vs_varient.c [deleted file]

index 2be4509b7a5ccdbd82f77c33e445f3c847c933b0..d10f9cd49cb93f86a8b02d275e887b64ddf92a15 100644 (file)
@@ -44,7 +44,7 @@ C_SOURCES = \
        draw/draw_vs_exec.c \
        draw/draw_vs_ppc.c \
        draw/draw_vs_sse.c \
-       draw/draw_vs_varient.c \
+       draw/draw_vs_variant.c \
        indices/u_indices_gen.c \
        indices/u_unfilled_gen.c \
        os/os_misc.c \
index 96ca566ff669f99f8b6a274b3f3e1794ed09bb0a..457911d2f1fbba4a56e2f9df1816b05ebf5fc205 100644 (file)
@@ -88,7 +88,7 @@ source = [
     'draw/draw_vs_exec.c',
     'draw/draw_vs_ppc.c',
     'draw/draw_vs_sse.c',
-    'draw/draw_vs_varient.c',
+    'draw/draw_vs_variant.c',
     #'indices/u_indices.c',
     #'indices/u_unfilled_indices.c',
     'indices/u_indices_gen.c',
diff --git a/src/gallium/auxiliary/draw/draw_vs_variant.c b/src/gallium/auxiliary/draw/draw_vs_variant.c
new file mode 100644 (file)
index 0000000..d8f030f
--- /dev/null
@@ -0,0 +1,341 @@
+/**************************************************************************
+ * 
+ * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas.
+ * All Rights Reserved.
+ * 
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ * 
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ * 
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
+ * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
+ * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ * 
+ **************************************************************************/
+
+ /*
+  * Authors:
+  *   Keith Whitwell <keith@tungstengraphics.com>
+  */
+
+#include "util/u_memory.h"
+#include "util/u_math.h"
+#include "draw/draw_context.h"
+#include "draw/draw_private.h"
+#include "draw/draw_vbuf.h"
+#include "draw/draw_vertex.h"
+#include "draw/draw_vs.h"
+#include "translate/translate.h"
+
+/* A first pass at incorporating vertex fetch/emit functionality into 
+ */
+struct draw_vs_variant_generic {
+   struct draw_vs_variant base;
+
+   struct draw_vertex_shader *shader;
+   struct draw_context *draw;
+   
+   /* Basic plan is to run these two translate functions before/after
+    * the vertex shader's existing run_linear() routine to simulate
+    * the inclusion of this functionality into the shader...  
+    * 
+    * Next will look at actually including it.
+    */
+   struct translate *fetch;
+   struct translate *emit;
+
+   unsigned temp_vertex_stride;
+};
+
+
+
+
+
+static void vsvg_set_buffer( struct draw_vs_variant *variant,
+                             unsigned buffer,
+                             const void *ptr,
+                             unsigned stride,
+                             unsigned max_index )
+{
+   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
+
+   vsvg->fetch->set_buffer(vsvg->fetch, 
+                           buffer, 
+                           ptr, 
+                           stride,
+                           max_index );
+}
+
+
+/* Mainly for debug at this stage:
+ */
+static void do_rhw_viewport( struct draw_vs_variant_generic *vsvg,
+                             unsigned count,
+                             void *output_buffer )
+{
+   char *ptr = (char *)output_buffer;
+   const float *scale = vsvg->base.vs->draw->viewport.scale;
+   const float *trans = vsvg->base.vs->draw->viewport.translate;
+   unsigned stride = vsvg->temp_vertex_stride;
+   unsigned j;
+
+   ptr += vsvg->base.vs->position_output * 4 * sizeof(float);
+
+   for (j = 0; j < count; j++, ptr += stride) {
+      float *data = (float *)ptr;
+      float w = 1.0f / data[3];
+
+      data[0] = data[0] * w * scale[0] + trans[0];
+      data[1] = data[1] * w * scale[1] + trans[1];
+      data[2] = data[2] * w * scale[2] + trans[2];
+      data[3] = w;
+   }
+}
+
+static void do_viewport( struct draw_vs_variant_generic *vsvg,
+                         unsigned count,
+                         void *output_buffer )
+{
+   char *ptr = (char *)output_buffer;
+   const float *scale = vsvg->base.vs->draw->viewport.scale;
+   const float *trans = vsvg->base.vs->draw->viewport.translate;
+   unsigned stride = vsvg->temp_vertex_stride;
+   unsigned j;
+
+   ptr += vsvg->base.vs->position_output * 4 * sizeof(float);
+
+   for (j = 0; j < count; j++, ptr += stride) {
+      float *data = (float *)ptr;
+
+      data[0] = data[0] * scale[0] + trans[0];
+      data[1] = data[1] * scale[1] + trans[1];
+      data[2] = data[2] * scale[2] + trans[2];
+   }
+}
+                         
+
+static void PIPE_CDECL vsvg_run_elts( struct draw_vs_variant *variant,
+                                      const unsigned *elts,
+                                      unsigned count,
+                                      void *output_buffer)
+{
+   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
+   unsigned temp_vertex_stride = vsvg->temp_vertex_stride;
+   void *temp_buffer = MALLOC( align(count,4) * temp_vertex_stride );
+   
+   if (0) debug_printf("%s %d \n", __FUNCTION__,  count);
+                       
+   /* Want to do this in small batches for cache locality?
+    */
+   
+   vsvg->fetch->run_elts( vsvg->fetch, 
+                          elts,
+                          count,
+                          vsvg->draw->instance_id,
+                          temp_buffer );
+
+   vsvg->base.vs->run_linear( vsvg->base.vs, 
+                              temp_buffer,
+                              temp_buffer,
+                              vsvg->base.vs->draw->pt.user.vs_constants,
+                              vsvg->base.vs->draw->pt.user.vs_constants_size,
+                              count,
+                              temp_vertex_stride, 
+                              temp_vertex_stride);
+
+   /* FIXME: geometry shading? */
+
+   if (vsvg->base.key.clip) {
+      /* not really handling clipping, just do the rhw so we can
+       * see the results...
+       */
+      do_rhw_viewport( vsvg,
+                       count,
+                       temp_buffer );
+   }
+   else if (vsvg->base.key.viewport) {
+      do_viewport( vsvg,
+                   count,
+                   temp_buffer );
+   }
+
+
+   vsvg->emit->set_buffer( vsvg->emit,
+                           0, 
+                           temp_buffer,
+                           temp_vertex_stride,
+                           ~0 );
+
+   vsvg->emit->set_buffer( vsvg->emit, 
+                           1,
+                           &vsvg->draw->rasterizer->point_size,
+                           0,
+                           ~0 );
+
+   vsvg->emit->run( vsvg->emit,
+                    0, count,
+                    vsvg->draw->instance_id,
+                    output_buffer );
+
+   FREE(temp_buffer);
+}
+
+
+static void PIPE_CDECL vsvg_run_linear( struct draw_vs_variant *variant,
+                                        unsigned start,
+                                        unsigned count,
+                                        void *output_buffer )
+{
+   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
+   unsigned temp_vertex_stride = vsvg->temp_vertex_stride;
+   void *temp_buffer = MALLOC( align(count,4) * temp_vertex_stride );
+       
+   if (0) debug_printf("%s %d %d (sz %d, %d)\n", __FUNCTION__, start, count,
+                       vsvg->base.key.output_stride,
+                       temp_vertex_stride);
+
+   vsvg->fetch->run( vsvg->fetch, 
+                     start,
+                     count,
+                     vsvg->draw->instance_id,
+                     temp_buffer );
+
+   vsvg->base.vs->run_linear( vsvg->base.vs, 
+                              temp_buffer,
+                              temp_buffer,
+                              vsvg->base.vs->draw->pt.user.vs_constants,
+                              vsvg->base.vs->draw->pt.user.vs_constants_size,
+                              count,
+                              temp_vertex_stride, 
+                              temp_vertex_stride);
+
+   if (vsvg->base.key.clip) {
+      /* not really handling clipping, just do the rhw so we can
+       * see the results...
+       */
+      do_rhw_viewport( vsvg,
+                       count,
+                       temp_buffer );
+   }
+   else if (vsvg->base.key.viewport) {
+      do_viewport( vsvg,
+                   count,
+                   temp_buffer );
+   }
+
+   vsvg->emit->set_buffer( vsvg->emit,
+                           0, 
+                           temp_buffer,
+                           temp_vertex_stride,
+                           ~0 );
+   
+   vsvg->emit->set_buffer( vsvg->emit, 
+                           1,
+                           &vsvg->draw->rasterizer->point_size,
+                           0,
+                           ~0 );
+   
+   vsvg->emit->run( vsvg->emit,
+                    0, count,
+                    vsvg->draw->instance_id,
+                    output_buffer );
+
+   FREE(temp_buffer);
+}
+
+
+
+
+
+static void vsvg_destroy( struct draw_vs_variant *variant )
+{
+   FREE(variant);
+}
+
+
+struct draw_vs_variant *
+draw_vs_create_variant_generic( struct draw_vertex_shader *vs,
+                                const struct draw_vs_variant_key *key )
+{
+   unsigned i;
+   struct translate_key fetch, emit;
+
+   struct draw_vs_variant_generic *vsvg = CALLOC_STRUCT( draw_vs_variant_generic );
+   if (vsvg == NULL)
+      return NULL;
+
+   vsvg->base.key = *key;
+   vsvg->base.vs = vs;
+   vsvg->base.set_buffer    = vsvg_set_buffer;
+   vsvg->base.run_elts      = vsvg_run_elts;
+   vsvg->base.run_linear    = vsvg_run_linear;
+   vsvg->base.destroy       = vsvg_destroy;
+
+   vsvg->draw = vs->draw;
+
+   vsvg->temp_vertex_stride = MAX2(key->nr_inputs,
+                                   vsvg->base.vs->info.num_outputs) * 4 * sizeof(float);
+
+   /* Build free-standing fetch and emit functions:
+    */
+   fetch.nr_elements = key->nr_inputs;
+   fetch.output_stride = vsvg->temp_vertex_stride;
+   for (i = 0; i < key->nr_inputs; i++) {
+      fetch.element[i].type = TRANSLATE_ELEMENT_NORMAL;
+      fetch.element[i].input_format = key->element[i].in.format;
+      fetch.element[i].input_buffer = key->element[i].in.buffer;
+      fetch.element[i].input_offset = key->element[i].in.offset;
+      fetch.element[i].instance_divisor = 0;
+      fetch.element[i].output_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
+      fetch.element[i].output_offset = i * 4 * sizeof(float);
+      assert(fetch.element[i].output_offset < fetch.output_stride);
+   }
+
+
+   emit.nr_elements = key->nr_outputs;
+   emit.output_stride = key->output_stride;
+   for (i = 0; i < key->nr_outputs; i++) {
+      if (key->element[i].out.format != EMIT_1F_PSIZE)
+      {      
+         emit.element[i].type = TRANSLATE_ELEMENT_NORMAL;
+         emit.element[i].input_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
+         emit.element[i].input_buffer = 0;
+         emit.element[i].input_offset = key->element[i].out.vs_output * 4 * sizeof(float);
+         emit.element[i].instance_divisor = 0;
+         emit.element[i].output_format = draw_translate_vinfo_format(key->element[i].out.format);
+         emit.element[i].output_offset = key->element[i].out.offset;
+         assert(emit.element[i].input_offset <= fetch.output_stride);
+      }
+      else {
+         emit.element[i].type = TRANSLATE_ELEMENT_NORMAL;
+         emit.element[i].input_format = PIPE_FORMAT_R32_FLOAT;
+         emit.element[i].input_buffer = 1;
+         emit.element[i].input_offset = 0;
+         emit.element[i].instance_divisor = 0;
+         emit.element[i].output_format = PIPE_FORMAT_R32_FLOAT;
+         emit.element[i].output_offset = key->element[i].out.offset;
+      }
+   }
+
+   vsvg->fetch = draw_vs_get_fetch( vs->draw, &fetch );
+   vsvg->emit = draw_vs_get_emit( vs->draw, &emit );
+
+   return &vsvg->base;
+}
+
+
+
+
+
diff --git a/src/gallium/auxiliary/draw/draw_vs_varient.c b/src/gallium/auxiliary/draw/draw_vs_varient.c
deleted file mode 100644 (file)
index d8f030f..0000000
+++ /dev/null
@@ -1,341 +0,0 @@
-/**************************************************************************
- * 
- * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas.
- * All Rights Reserved.
- * 
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the
- * "Software"), to deal in the Software without restriction, including
- * without limitation the rights to use, copy, modify, merge, publish,
- * distribute, sub license, and/or sell copies of the Software, and to
- * permit persons to whom the Software is furnished to do so, subject to
- * the following conditions:
- * 
- * The above copyright notice and this permission notice (including the
- * next paragraph) shall be included in all copies or substantial portions
- * of the Software.
- * 
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
- * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
- * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
- * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
- * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
- * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
- * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
- * 
- **************************************************************************/
-
- /*
-  * Authors:
-  *   Keith Whitwell <keith@tungstengraphics.com>
-  */
-
-#include "util/u_memory.h"
-#include "util/u_math.h"
-#include "draw/draw_context.h"
-#include "draw/draw_private.h"
-#include "draw/draw_vbuf.h"
-#include "draw/draw_vertex.h"
-#include "draw/draw_vs.h"
-#include "translate/translate.h"
-
-/* A first pass at incorporating vertex fetch/emit functionality into 
- */
-struct draw_vs_variant_generic {
-   struct draw_vs_variant base;
-
-   struct draw_vertex_shader *shader;
-   struct draw_context *draw;
-   
-   /* Basic plan is to run these two translate functions before/after
-    * the vertex shader's existing run_linear() routine to simulate
-    * the inclusion of this functionality into the shader...  
-    * 
-    * Next will look at actually including it.
-    */
-   struct translate *fetch;
-   struct translate *emit;
-
-   unsigned temp_vertex_stride;
-};
-
-
-
-
-
-static void vsvg_set_buffer( struct draw_vs_variant *variant,
-                             unsigned buffer,
-                             const void *ptr,
-                             unsigned stride,
-                             unsigned max_index )
-{
-   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
-
-   vsvg->fetch->set_buffer(vsvg->fetch, 
-                           buffer, 
-                           ptr, 
-                           stride,
-                           max_index );
-}
-
-
-/* Mainly for debug at this stage:
- */
-static void do_rhw_viewport( struct draw_vs_variant_generic *vsvg,
-                             unsigned count,
-                             void *output_buffer )
-{
-   char *ptr = (char *)output_buffer;
-   const float *scale = vsvg->base.vs->draw->viewport.scale;
-   const float *trans = vsvg->base.vs->draw->viewport.translate;
-   unsigned stride = vsvg->temp_vertex_stride;
-   unsigned j;
-
-   ptr += vsvg->base.vs->position_output * 4 * sizeof(float);
-
-   for (j = 0; j < count; j++, ptr += stride) {
-      float *data = (float *)ptr;
-      float w = 1.0f / data[3];
-
-      data[0] = data[0] * w * scale[0] + trans[0];
-      data[1] = data[1] * w * scale[1] + trans[1];
-      data[2] = data[2] * w * scale[2] + trans[2];
-      data[3] = w;
-   }
-}
-
-static void do_viewport( struct draw_vs_variant_generic *vsvg,
-                         unsigned count,
-                         void *output_buffer )
-{
-   char *ptr = (char *)output_buffer;
-   const float *scale = vsvg->base.vs->draw->viewport.scale;
-   const float *trans = vsvg->base.vs->draw->viewport.translate;
-   unsigned stride = vsvg->temp_vertex_stride;
-   unsigned j;
-
-   ptr += vsvg->base.vs->position_output * 4 * sizeof(float);
-
-   for (j = 0; j < count; j++, ptr += stride) {
-      float *data = (float *)ptr;
-
-      data[0] = data[0] * scale[0] + trans[0];
-      data[1] = data[1] * scale[1] + trans[1];
-      data[2] = data[2] * scale[2] + trans[2];
-   }
-}
-                         
-
-static void PIPE_CDECL vsvg_run_elts( struct draw_vs_variant *variant,
-                                      const unsigned *elts,
-                                      unsigned count,
-                                      void *output_buffer)
-{
-   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
-   unsigned temp_vertex_stride = vsvg->temp_vertex_stride;
-   void *temp_buffer = MALLOC( align(count,4) * temp_vertex_stride );
-   
-   if (0) debug_printf("%s %d \n", __FUNCTION__,  count);
-                       
-   /* Want to do this in small batches for cache locality?
-    */
-   
-   vsvg->fetch->run_elts( vsvg->fetch, 
-                          elts,
-                          count,
-                          vsvg->draw->instance_id,
-                          temp_buffer );
-
-   vsvg->base.vs->run_linear( vsvg->base.vs, 
-                              temp_buffer,
-                              temp_buffer,
-                              vsvg->base.vs->draw->pt.user.vs_constants,
-                              vsvg->base.vs->draw->pt.user.vs_constants_size,
-                              count,
-                              temp_vertex_stride, 
-                              temp_vertex_stride);
-
-   /* FIXME: geometry shading? */
-
-   if (vsvg->base.key.clip) {
-      /* not really handling clipping, just do the rhw so we can
-       * see the results...
-       */
-      do_rhw_viewport( vsvg,
-                       count,
-                       temp_buffer );
-   }
-   else if (vsvg->base.key.viewport) {
-      do_viewport( vsvg,
-                   count,
-                   temp_buffer );
-   }
-
-
-   vsvg->emit->set_buffer( vsvg->emit,
-                           0, 
-                           temp_buffer,
-                           temp_vertex_stride,
-                           ~0 );
-
-   vsvg->emit->set_buffer( vsvg->emit, 
-                           1,
-                           &vsvg->draw->rasterizer->point_size,
-                           0,
-                           ~0 );
-
-   vsvg->emit->run( vsvg->emit,
-                    0, count,
-                    vsvg->draw->instance_id,
-                    output_buffer );
-
-   FREE(temp_buffer);
-}
-
-
-static void PIPE_CDECL vsvg_run_linear( struct draw_vs_variant *variant,
-                                        unsigned start,
-                                        unsigned count,
-                                        void *output_buffer )
-{
-   struct draw_vs_variant_generic *vsvg = (struct draw_vs_variant_generic *)variant;
-   unsigned temp_vertex_stride = vsvg->temp_vertex_stride;
-   void *temp_buffer = MALLOC( align(count,4) * temp_vertex_stride );
-       
-   if (0) debug_printf("%s %d %d (sz %d, %d)\n", __FUNCTION__, start, count,
-                       vsvg->base.key.output_stride,
-                       temp_vertex_stride);
-
-   vsvg->fetch->run( vsvg->fetch, 
-                     start,
-                     count,
-                     vsvg->draw->instance_id,
-                     temp_buffer );
-
-   vsvg->base.vs->run_linear( vsvg->base.vs, 
-                              temp_buffer,
-                              temp_buffer,
-                              vsvg->base.vs->draw->pt.user.vs_constants,
-                              vsvg->base.vs->draw->pt.user.vs_constants_size,
-                              count,
-                              temp_vertex_stride, 
-                              temp_vertex_stride);
-
-   if (vsvg->base.key.clip) {
-      /* not really handling clipping, just do the rhw so we can
-       * see the results...
-       */
-      do_rhw_viewport( vsvg,
-                       count,
-                       temp_buffer );
-   }
-   else if (vsvg->base.key.viewport) {
-      do_viewport( vsvg,
-                   count,
-                   temp_buffer );
-   }
-
-   vsvg->emit->set_buffer( vsvg->emit,
-                           0, 
-                           temp_buffer,
-                           temp_vertex_stride,
-                           ~0 );
-   
-   vsvg->emit->set_buffer( vsvg->emit, 
-                           1,
-                           &vsvg->draw->rasterizer->point_size,
-                           0,
-                           ~0 );
-   
-   vsvg->emit->run( vsvg->emit,
-                    0, count,
-                    vsvg->draw->instance_id,
-                    output_buffer );
-
-   FREE(temp_buffer);
-}
-
-
-
-
-
-static void vsvg_destroy( struct draw_vs_variant *variant )
-{
-   FREE(variant);
-}
-
-
-struct draw_vs_variant *
-draw_vs_create_variant_generic( struct draw_vertex_shader *vs,
-                                const struct draw_vs_variant_key *key )
-{
-   unsigned i;
-   struct translate_key fetch, emit;
-
-   struct draw_vs_variant_generic *vsvg = CALLOC_STRUCT( draw_vs_variant_generic );
-   if (vsvg == NULL)
-      return NULL;
-
-   vsvg->base.key = *key;
-   vsvg->base.vs = vs;
-   vsvg->base.set_buffer    = vsvg_set_buffer;
-   vsvg->base.run_elts      = vsvg_run_elts;
-   vsvg->base.run_linear    = vsvg_run_linear;
-   vsvg->base.destroy       = vsvg_destroy;
-
-   vsvg->draw = vs->draw;
-
-   vsvg->temp_vertex_stride = MAX2(key->nr_inputs,
-                                   vsvg->base.vs->info.num_outputs) * 4 * sizeof(float);
-
-   /* Build free-standing fetch and emit functions:
-    */
-   fetch.nr_elements = key->nr_inputs;
-   fetch.output_stride = vsvg->temp_vertex_stride;
-   for (i = 0; i < key->nr_inputs; i++) {
-      fetch.element[i].type = TRANSLATE_ELEMENT_NORMAL;
-      fetch.element[i].input_format = key->element[i].in.format;
-      fetch.element[i].input_buffer = key->element[i].in.buffer;
-      fetch.element[i].input_offset = key->element[i].in.offset;
-      fetch.element[i].instance_divisor = 0;
-      fetch.element[i].output_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
-      fetch.element[i].output_offset = i * 4 * sizeof(float);
-      assert(fetch.element[i].output_offset < fetch.output_stride);
-   }
-
-
-   emit.nr_elements = key->nr_outputs;
-   emit.output_stride = key->output_stride;
-   for (i = 0; i < key->nr_outputs; i++) {
-      if (key->element[i].out.format != EMIT_1F_PSIZE)
-      {      
-         emit.element[i].type = TRANSLATE_ELEMENT_NORMAL;
-         emit.element[i].input_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
-         emit.element[i].input_buffer = 0;
-         emit.element[i].input_offset = key->element[i].out.vs_output * 4 * sizeof(float);
-         emit.element[i].instance_divisor = 0;
-         emit.element[i].output_format = draw_translate_vinfo_format(key->element[i].out.format);
-         emit.element[i].output_offset = key->element[i].out.offset;
-         assert(emit.element[i].input_offset <= fetch.output_stride);
-      }
-      else {
-         emit.element[i].type = TRANSLATE_ELEMENT_NORMAL;
-         emit.element[i].input_format = PIPE_FORMAT_R32_FLOAT;
-         emit.element[i].input_buffer = 1;
-         emit.element[i].input_offset = 0;
-         emit.element[i].instance_divisor = 0;
-         emit.element[i].output_format = PIPE_FORMAT_R32_FLOAT;
-         emit.element[i].output_offset = key->element[i].out.offset;
-      }
-   }
-
-   vsvg->fetch = draw_vs_get_fetch( vs->draw, &fetch );
-   vsvg->emit = draw_vs_get_emit( vs->draw, &emit );
-
-   return &vsvg->base;
-}
-
-
-
-
-